Also: IPv4, IPv6, IP address, private address, public address
IP addressing is the scheme that gives every interface on a network a numeric address, with IPv4's four-byte addresses now supplemented by IPv6's much larger space, and with private ranges reserved for use inside networks that reach the internet through translation.
Our take. IPv4 ran out of addresses years ago and the industry has spent the time since building elaborate workarounds rather than finishing the migration. Learn IPv6 properly now, because every cloud provider and every mobile network already runs it, and the engineer who treats it as optional will be the one confused by an address with colons in a production incident.
An IPv4 address is four numbers, each from zero to two hundred and fifty-five, written with dots. Part of it identifies the network and the rest identifies the host within it, and where the split falls is the subject of subnetting. Certain ranges are reserved for private use inside organisations and never appear on the public internet; a home router, a corporate network and a cloud virtual network all use them, and the fact that millions of networks share the same private ranges is only workable because translation at the edge rewrites them on the way out. Public addresses are scarce, allocated by regional registries, and the reason the internet has any room left is that scarcity.
Private ranges: three blocks set aside for internal use, reachable from the internet only through translation. The same ranges are reused in every organisation on earth.
Loopback: an address that always means this machine, used for testing and for services that should never be reachable from outside.
Link-local: addresses a device gives itself when nothing else assigned one, which is what an interface shows when the address server is unreachable.
Multicast and broadcast: addresses that reach many hosts at once; broadcast is how a device finds an address server before it has an address.
Documentation ranges: blocks reserved for examples in books and exams so that nobody's real network appears in a textbook.
IPv6 addresses are eight groups of four hexadecimal digits, written with colons and shortened by dropping leading zeros and one run of empty groups. The space is large enough that each device may carry a public address and translation becomes unnecessary, which is the design's point, and the engineering habits that IPv4 scarcity forced, translation, tiny subnets, address hoarding, are the ones IPv6 removes. It also changes the exams. The networking certifications test both families and expect you to shorten an IPv6 address correctly and to know which kinds of address a host holds at once; the cloud exams assume you can read either in a security rule.
In practice
A laptop on an office network shows an address in a private range, and the same address exists in ten thousand other offices. When it fetches a web page, the office router replaces that private address with the office's single public one on the way out and reverses the swap on the way back, keeping a table of which internal device asked for what. The web server sees the office, not the laptop. The same laptop on a modern mobile network holds a public IPv6 address of its own, no translation happens, and the server sees the device directly. Both work; only the second is how the design was meant to be used.
Reading an address: the prefix length after the slash says how many leading bits are the network. The rest is the host.
Recognising a private address: three IPv4 blocks and one IPv6 range mean internal. Seeing one on the public internet means something is misconfigured or spoofed.
Spotting a failed assignment: a link-local address on an interface that should have a real one means the address server was not reached.
Addressing is the scheme; subnetting is deciding where the network part of an address ends and the host part begins, and carving a block into smaller networks accordingly.
Network address translation is the mechanism that lets private addresses reach the public internet by rewriting them at the edge. Addressing defines the ranges; translation is the workaround for their scarcity.
DNS turns a name into an address. The address is what packets are sent to; the name is what people remember, and the two are looked up, not the same thing.
Key takeaways
→An address splits into network and host; the prefix length says where.
→Private ranges are reused everywhere and reach the internet only through translation.
→IPv6 removes the scarcity and the workarounds; learn it as a first-class family.
Certifications that test this
Vendor exams whose syllabus covers this concept — facts, cost and a preparation path on each page.
A rotating selection from the course directory, drawn from the subcategories where this concept is taught rather than picked for it. Details, price and the provider link are on the course page.
.NET is a modern, innovative, open source development platform and developers love it. NET best suits businesses that l…
Udemy
FAQ
Is an IP address an identity?
No. It identifies an interface on a network at a moment, it is often shared through translation, and it changes when a device moves. Treating an address as a person, in logs or in access rules, is a common and consequential mistake.
Do I still need to learn IPv4?
Yes, for years yet. Most internal networks and a great deal of the public internet still run it, the exams weight it heavily, and the mental model of network and host parts transfers directly to IPv6 once learned.
Why are IPv6 addresses so long?
Because the space is enormous by design, large enough that every device, sensor and virtual machine can hold a public address without anyone rationing them. The length is the price of never running out again, and the shortening rules make them manageable in practice.
Sources
The primary text this definition rests on. Read it before you trust ours.
